- AI TALK
Pi 6, via Indonesia 23 • June 15th, 2022 • 7:00 PM CET
Augmenting Human Intuition
Augmenting Human Intuition
Learn from Petar Veličković, author of "Advancing mathematics by guiding human intuition with AI", how machines empower researchers conjecturing new approaches to long-standing open problems.
Petar Veličković
Staff Research Scientist at DeepMindFor the past few years, Petar Veličković has been working on a challenging project: teaching machines to assist humans with proving difficult theorems and conjecturing new approaches to long-standing open problems. His team has demonstrated that analyzing and interpreting the outputs of (graph) neural networks is a concrete solution. It also independently derived novel top-tier mathematical results in areas as diverse as representation theory and knot theory.
During the event, Petar will share his findings from a personal perspective, with key details of his team’s modeling work – already featured by Nature
OUR GUEST
About
Petar Veličković
Staff research scientist at DeepMind, affiliated Lecturer at the University of Cambridge, and an associate of Clare Hall, Cambridge.
His research concerns geometric deep learning — devising neural network architectures that respect the invariances and symmetries in data. Within this area, he focus on graph representation learning and its applications in algorithmic reasoning and computational biology.
Petar is the first author of Graph Attention Networks — a popular convolutional layer for graphs — and Deep Graph Infomax — a popular self-supervised learning pipeline for graphs. His research has also been used in substantially improving the travel-time predictions in Google Maps.
6 Takeaways You'll Get
From Attending This Event
Making Transformers memory efficient without sacrificing accuracy
Fine-tuning state-of-the-art models without datacenter-scale hardware resources
Adapting Transformers to run over 1 million of tokens on a single GPU or TPU device
Replacing dot-product attention by one that uses locality-sensitive hashing
Using reversible residual layers to store activations only once in the training process
Applying Transformers efficiency techniques to new use case scenarios
Event Agenda
-
Welcome to Pi Campus 7:00 pm
Marco Trombetti, Pi Campus founder and CEO
-
Introduction 7:05 pm
Simone Scardapane, tenure-track assistant professor at Sapienza University of Rome.
-
Keynote 7:10 pm
Petar Veličković, Staff Research Scientist at DeepMind
-
Q&A 7:55 pm
Participants will have the opportunity to ask questions related to the talk.
-
Networking Aperitif 8:30 pm
The event will end with some networking time.
Get Invited
These events are reserved to Pi Campus founders and our guests. If you were not invited, you can request an invite here. Be aware that requesting an invitation does not guarantee an invitation from us.